other characteristics now associated with polygraph testing. For example, any other technique that potential examinees believed to be valid for detecting deception would be likely to elicit admissions and to have deterrent effects. Even a technique that examinees believed to be invalid but that they also believed would be treated as valid might bring some benefits of deterrence.
The term validity, when applied to polygraph testing, normally refers only to the psychophysiological test. That is, the polygraph is said to be valid only if deception is strongly and uniquely associated with a discernable pattern in the record of physiological responses made on or from the polygraph. Chapter 3 discusses the scientific basis for believing that deception produces specific psychological and physiological processes that influence polygraph readings, which indicates the construct validity of the polygraph test. Chapters 4 and 5 discuss the evidence on the empirical association between deception and polygraph test results, which indicates the accuracy of the polygraph. As noted above, the utility of the polygraph depends on more than just its validity, but these utility aspects should be considered separately from the issue of validity of the instrument for measuring deception.
It is possible to assess the ability of interviewers to detect deception from visible and audible cues in the interview, generally referred to as demeanor (e.g., facial expression, posture, voice quality). Considerable scientific effort has been devoted to the development of techniques for detecting deception from such behavioral indicators. Although this is not a major focus of the present study, we discuss the research evidence on this detection briefly in Chapter 6. In the context of evaluating the validity of the polygraph instrument, it is necessary to exclude the possibility that the examiner’s judgment was affected in part by the examinee’s demeanor, since that is not what the polygraph instrument measures. This can be done in part by evaluating the performance of polygraph assessments made by computer programs or by trained examiners who have access only to the record of physiological responses. If examiners who actually interviewed the subject make the same judgments as result from these other methods, their assessments are unlikely to have been affected directly by the examinee’s demeanor.12